Digi-Key has changed and it is not very good

[rsjsouza]

This is the reason websites continually change. It's not just "change for the sake of change".

Although not 100% of the cases, in my experience that happens enough times to be very significant. I have been in many UI/website/platform decision making committees and, in my experience, the bad decisions are made based on several factors: either "pie-in-the-sky" goals (we need to become an Amazon or a Facebook), using weak arguments (the ever ethereal "market trend" or modern "look and feel"), feedback shutdown due to bad planning/timing (we like your suggestion but can't do anything as the deadline is too close) or based on very small sample size "research" and/or "studies" (only a very small fraction of customers take their time to answer usability and feedback questionnaires, let alone the ones that subject themselves to early user adoption test). In my experience, they are incredibly correlated with internal management change - the origin of my original comment.

The great number of good decisions exist, but they are usually based on solid research and feedback - yes, this is one of the few cases I dare to say that "design by committee" tends to have more successful stories than "single ruler party". 

A good example of this is the recent picklist change, where the picklists went from normal html select lists, to the (imo) frustrating picklists that no longer perform like a normal html select list. Using CTRL and SHIFT to select multiple is items is broken now, and replaced by selecting individual items by clicking on them. This change is frustrating, but I understand why they did it. I've stood behind at least 3 junior EE's who while using digikey, would select one item from a picklist, view the results, then go back, select another, view those results. When I pointed out that they could select multiple items, they were surprised that this (basic, default html select) feature existed. I observed these same people trying to select multiple items by just clicking on them (like you can do now). This is why websites like DigiKey make these changes, because while you may know how to use the tool very efficiently, there's bound to be lots of people less experienced learning or struggling.

I don't quite follow your example. Despite the change to something (in your opinion) worse, you showed the junior EEs how to do a multiple selection and this brought one more tool to their belt. The way I read it, the operative word on your example is "learning".

I personally don't have a beef with "sticky-click" picklists, as they are one of the most usable ways if you are on a mobile platform without a keyboard. Besides, you can still use Shift to select ranges and, even better, you can select non-contiguous ranges using the same key. However, intuitiveness is not the holy grail and certain non-intuitive things need to be inserted to make people pause before taking a decision - the "OK" or "Apply" buttons in configuration screens, for example, which are disappearing from newer UIs.

It's a tricky balance to keep. On the one hand you don't want to alienate your existing customers, but on the other hand you want to make it as easy as possible for new users to make their first order. Inevitably the balance usually tips more towards the new users, as experienced users are probably less likely to leave over UI changes. I know I won't leave over the new picklists. While I really don't like the new picklists, I know that the value DigiKey offers is in more than just their picklists.

I agree with the balance part of it. However, I also agree with the sentiment that UIs can make people leave your platform - see Arrow.com, for example. For a B2B e-commerce, however, the issue must be much worse than a few clicks here and there (automation is usually much more valued).

[floobydust]

This is about mass incompetence with their new website. The basics of project management- software requirements, signoff, coding, testing - all missing.
I'd roll some heads because it's obvious nobody knows what an LED is or even owns a soldering iron over there lol.

Example: looking for a voltage range in a component group.
Can we enter a range in "Search Filter" ? NO. "2-3" gives nothing.
Can't do it at all with MCU's, if looking for a 5V or 1.8V part. This has been complained about by many and nothing done.

For LED's. let's try type in a friendly "4" for Vf.
Oh look, it's our old friend of inept website coding - Radix sort, great with punched cards in 1923, as if numeric values are text.
Results (in pic) include any entry containing a "4", up to 48V or 64V. Type in "2.9" and 12.9 is a hit 
TOTALLY STUPID.

Most corporations would now look at (and fix) all parameter filters, but no we get to die a slow death complaining here byte by byte. Or go to Mouser lol.

[james_s]

Seriously what kind of attitude is this? Digikey doesn't exist to rant at the clouds about youngsters these days. They're here to attract new customers and keep existing ones and grow their business. As is every business. Only they know whether their changes accomplish this goal or not, because they're the only ones with relevant data. I hope they fire any web designer that can't set up simple A/B tests without going off on a tangential rant about how everyone is stupid nowadays.

Your attitude is to change something making it more difficult for one group in an attempt to make it easier for another group, how is that any better? This is such a trivial thing that simply informing someone how to do it would solve the "problem". Perhaps you'll "get it" once you're old enough to have gotten bored of change for the sake of change and tired of re-leaning things you already spent time learning and having your workflow and muscle memory disrupted by pointless changes.. Get off your high horse, you're not any more virtuous than the rest of us and you're not going to be fundamentally different when you grow up than the people you're complaining about..

[james_s]

For LED's. let's try type in a friendly "4" for Vf.
Oh look, it's our old friend of inept website coding - Radix sort, great with punched cards in 1923, as if numeric values are text.
Results (in pic) include any entry containing a "4", up to 48V or 64V. Type in "2.9" and 12.9 is a hit 
TOTALLY STUPID.

Most corporations would now look at (and fix) all parameter filters, but no we get to die a slow death complaining here byte by byte. Or go to Mouser lol.

I did just that in my most recent order, I got so frustrated trying to find what I was looking for that I closed the browser in disgust and ordered it from Mouser instead. Not that Mouser's website is any kind of gold standard but at least the sorting of stuff like voltages isn't completely broken and useless. 

This is the stuff that really makes a difference. I can adjust to a change in the way the picklists work if the values are at least sorted properly so I can select a range, but when it's just completely broken, I'm sorry but this train wreck is a display of gross incompetence and is simply inexcusable.

[SVFeingold]

This is such a trivial thing that simply informing someone how to do it would solve the "problem".

I agree, but they could probably say the same about the new system.   

Perhaps you'll "get it" once you're old enough to have gotten bored of change for the sake of change and tired of re-leaning things you already spent time learning and having your workflow and muscle memory disrupted by pointless changes.. Get off your high horse, you're not any more virtuous than the rest of us and you're not going to be fundamentally different when you grow up than the people you're complaining about..

"It's inevitable, you'll be bitter when you're older too." Sorry, I don't subscribe. And anyway I agree with you about change for the sake of change. I've got plenty of examples I could go on about. It's not that I don't "get it," it's that there's no reason to condition myself to view every issue in the light of "how does this make me smart and everyone else a moron." There's your high horse right next to mine. Like the classic dad move where every conversation, no matter how it starts, somehow ends up about being WWII. And refusing to change is just as bad. Ask basically every convoluted, archaic ECAD/MCAD/FEA software out there that requires 4 keystrokes and a command line argument to do some common function that could easily be done with a single click. It's a long list to choose from. But don't you dare change a single icon or the hordes of engineers who think UI design peaked at Windows 3.1 will raise hell. Shit, now my high horse is getting taller.

To make it clear I agree with you that there was nothing really wrong with the old picklist. And I agree with nearly ever complaint in this thread. Especially flooby about the terrible sorting, but he managed to do it without taking it in a "kids these days" direction. 

[tooki]

That's changing now, thanks to heat pump dryers, which are vastly more energy efficient than traditional vented dryers. (And they're gentler on the clothes, since they don't heat the clothes up as hot.)

My friend got one of those recently, he likes it, they're still very rare in the US though, he's the only person I know who has one. I'd heard of them before though, apparently they've been fairly common in Europe for a while.

Correct. European energy standards have made non-heat-pump dryers essentially illegal to sell, IIRC. The apartment I live in came with one, and I've been here for over 7 years. I've been exceptionally happy with it, as the lower temperature means I can put things in the dryer without fear of shrinking them.

A friend of mine in USA got one about 2 years ago. But yeah, we had to look around to even find one. (He ended up going with Miele.)

[tooki]

They might save a lot of electricity.  Would be interesting to see the numbers.

The numbers I've seen say that they use 50-60% as much energy as vented dryers, and only 35% as much energy as (non-heat-pump) condensing (ventless) dryers, which are hellishly inefficient.

[tooki]

I'm glad to see these viewpoints laid out so passionately. It really helps me identify the kind of engineer I don't want to become as I roll through my 30s and beyond.

Everyone is stupid if they don't use my workflow. New customers are confused by the UI? Good. They're dumb anyway. How'd they get to be engineers?  Anyone going through a 4 year program should know how to use this website! Good riddance! Digikey doesn't need such dumb customers.

Seriously what kind of attitude is this? Digikey doesn't exist to rant at the clouds about youngsters these days. They're here to attract new customers and keep existing ones and grow their business. As is every business. Only they know whether their changes accomplish this goal or not, because they're the only ones with relevant data. I hope they fire any web designer that can't set up simple A/B tests without going off on a tangential rant about how everyone is stupid nowadays.

As someone who is a) old enough to have established preferences and experienced things change for the sake of change, but b) has always had an interest in usability, c) has worked as a user interface designer, including at a usability agency where we did do user testing, d) has always hated the elitist attitude of Linux, etc, and d) has always had an interest in electronics, and is now changing to that as a career, I think you're absolutely right.

Now that said: yes, I agree that some of their design changes aren't great for my workflow. The picklist threw me off for a bit at first until I got used to it. Which took about 15 minutes. Do I like it more? Not sure yet, but there have been many times in the past when I accidentally let go of CTRL and wiped the selection I spent a while making. So I get the motivation. It's ironic to rant about dummies who don't understand a simple UI considering how it's arguably even simpler now. Glass houses, etc.

Yep. In all fairness, I'm not sure what'd be wrong with lists of checkboxes, since that's essentially what the new lists are. But with the old ones, it's so easy to accidentally wipe out a complex selection, just as you said.

What a lot of people forget is that a good interface buffers you from errors or lets you recover from them easily. It's not just beginners who make mistakes, everyone does, all the time. That's why "undo" was revolutionary. (And why it shocks me how, especially on mobile, Windows and Linux, there are still lots of apps with poor or nonexistent undo.)

This is the same attitude Linux fanboys have. I wonder how "everyone who can't use a command line is stupid" Linux is doing vs MacOS in the consumer market? I wonder why that could be.

Yup.

Doesn't mean everything DK does is perfect but they're not going to listen to people ranting about young engineers in lieu of constructive criticism. I'm as opinionated as anyone else but I try to consider things from other viewpoints in case I'm wrong. And frankly the more I progress in my career, the more and more I encounter mid-to-late career engineers who are needlessly bitter/angry about literally every little thing. It takes actual work to not go that route and folding your arms saying "everyone's dumb but me" isn't the way. Anecdotally this attitude has been much more prevalent in the EE field than any other field I work with. Who knows.

I see it in EE, but also with old-school IT/CS people (who needn't be old people, there are plenty of young ones who still like the old-school mindset). For a lot of engineers, it's really hard to step into the mindset of people who aren't identically technically-minded as them. I worded that that way because I've seen plenty of people who have extraordinary technical expertise in one area, but not another, and as such, you can have someone who is a gifted engineer in one area, but nonetheless is a relatively unskilled computer user.

Plus, why should stuff be needlessly complex to use? User interfaces need to be as simple as possible, but as complex as needed (but no more!). That's why it's acceptable for professional software to be more complex than consumer software. But if the needed pro functionality can be accomplished an easier way, then it should be.

I also have to remind people that the easiest interface is the one the user already knows. In other words, if you can make your interface work like something the user already knows, then it's the easiest of all. So even if you have an idea that works well in isolation, if it's very different from everything else, it is still likely to be harder to use in practice. Incremental change is a good way to overcome this, by slightly extending an existing interface bit by bit (for example, how someone came up with the "pull to refresh" action in mobile phone lists: it extends the existing interface seamlessly).

On-topic: When I was a new engineer I had no problems with the picklists (yay) but I admit that Digikey wasn't really useful to me. I wasn't working with an existing BOM, rather was looking for random parts to do cool projects. In other words: trying to discover new products. And it wasn't really great for that. It still isn't IMO. Nowadays I know exactly what I'm looking for so Digikey is much more usable for me, but I understand the struggle for new and/or inexperienced engineers. It took a while for Digikey to "make sense." Until that point I frequented Sparkfun and Adafruit much more often because they cater to new engineers and really just to exploring and discovering cool stuff. They're not really comparable. I definitely don't want Digikey to change that much but there's probably something they could borrow that would make it more enticing for new engineers. Throwing a bone to a new engineer today might mean tens of thousands of dollars in POs in the future.

Honestly, what's missing is old-school catalogs. I have never encountered a web shop that manages to recreate the browseability of a paper catalog. (DK tried: in addition to the parametric search and categories, there's a whole separate layer of content as the online catalog and the manufacturer pages, etc. But it never caught on.) I love perusing catalogs to discover things I didn't know to look for. Searches are great for when you know exactly what you need and what it's called. But it's terrible for just random discovery.

Your attitude is to change something making it more difficult for one group in an attempt to make it easier for another group, how is that any better? This is such a trivial thing that simply informing someone how to do it would solve the "problem". Perhaps you'll "get it" once you're old enough to have gotten bored of change for the sake of change and tired of re-leaning things you already spent time learning and having your workflow and muscle memory disrupted by pointless changes.. Get off your high horse, you're not any more virtuous than the rest of us and you're not going to be fundamentally different when you grow up than the people you're complaining about..

You're accusing him of being on a high horse?!?    Then what are you riding, an elephant?

A good design makes it better for everyone. "I had to learn to suffer with the old, hard interface, so so should you" is not a sensible argument against improvement. (But it's absolutely the attitude many espouse.) A truly exceptional interface is so transparent that you don't need to actively "learn" it, it just works the way you expect, without thought, letting you focus on your task rather than on the tool you're using to accomplish it. (This is what IMHO the Mac has always been far better at than anyone else: the interface just doesn't get in my way the way Windows does.) But any change is going to involve some adjustment for existing users, even if it's objectively a change for the better.

I actually agree that in the past few years, there's been a lot of pointless change in computer interfaces, a substantial amount of which is absolutely for the worse. (Like removing all visible cues to features? Hello, insanity! And the nonsense of designs so "airy" that there's less content on a 27" screen than on an old 14" 640x480 CRT back in 1994...) But that fact doesn't mean all change is bad! Some things were bad design decisions at the time, or were imposed by technical limitations that no longer apply. Such designs deserve to be replaced. 

[SilverSolder]

People generally hate UI changes, because they have to make an effort to (re)learn something they already knew how to do.

A friend in the financial industry had a spectacular example of this where they moved the Cancel and OK buttons a couple of centimeters to one side on a touch screen GUI for a trading system, to make room for a new feature. 

The customers went berserk and rang the help desk down all day...  -   it turns out they weren't even looking at the screens, they had the positions of all the controls embedded in tactile memory and were talking on the phone while their "fingers did the walking" on the touch screen!  They were hitting "Cancel" when they meant "OK"....    It was so bad, they had to issue an emergency release overnight to put the controls back in the exact same location as before the change.

[cortex_m0]

I've been shopping for components for a new design lately, and I actually quite like the new parametric search interface. It's very responsive. And I'm pleased not to need to filter out a bunch of equivalent parts with packaging options (e.g. DigiReel, Cut Tape, Full Reel) any longer. That was jarring at first, but it's a good change. I used to reflexively pick "Cut Tape" to eliminate duplicates, but that also eliminated parts that were sold in trays, tubes, etc.

I wouldn't mind if they took that a step further, and condensed parts which are offered in various packages by the manufacturer into those packaging listings, even if it is technically a different manufacturer part number. For my needs, doing small builds, it doesn't matter much if the part comes to me on a tray or on a Digi-Reel.

Someone here complained that you couldn't CTRL+Select or SHIFT+Select in the new parametric search. Its been working fine for me in Firefox. I was able to select all AVRs which will operate on 3.3V pretty easily using the traditional keystrokes.

However, it's unfortunate that the new parametric search took months after public rollout to get the obvious bugs resolved.

[SVFeingold]

IMO a big step forward would be figuring out how to bin features together. I mean look at this nonsense in the picture...

Instead of just choosing the protocols you're after from a short list and having it show you all relevant modules, every single configuration is a unique line item. This happens with most kinds of parts on there. LEDs were already mentioned earlier.

If I had to pick one thing to change it'd be that.

[Red Squirrel]

Yeah I find that annoying too.  Also they should really use a different type of element for numerical data.  Ex voltages.  Instead of having a list of voltages and having to check each one individually there should be a simple min/max option.    A lot of the time they have a bunch of arbitrary voltage ranges within the options. 

I think what happens is when they import data it probably just looks at the datasheets so like if an item is rated between 5-15 volts, rather than have that item show up if the user picks any of those values, it will actually create a value for 5-15.    So if you only pick 5v then you won't get that item show up because you also need to pick the explicit 5-15 option too.   At least that's my guess.

[SilverSolder]

They could probably ask manufacturers for help with that - after all, it is in the manufacturer's interests to have their products displayed correctly.  So their rep shouldn't have a problem filling in a spreadsheet for their products.

[asmi]

This is super-amusing fail    

[gussy]

This is super-amusing fail    

Thats' a good one

FWIW, whenever I see errors like that I use the "Report an Error" button and write a quick description of the problem. It takes less than 30 seconds and helps pay it forward to other users of the site. The errors I've reported seem to be resolved quickly.

[Red Squirrel]

I guess if you're passing high enough voltage through that cable it does not really matter what you use, so flash chips will do the trick. 

[james_s]

FWIW, whenever I see errors like that I use the "Report an Error" button and write a quick description of the problem. It takes less than 30 seconds and helps pay it forward to other users of the site. The errors I've reported seem to be resolved quickly.

When they start sending me a paycheck to be their QA engineer I'll start taking the time to file bug reports.

[gussy]

When they start sending me a paycheck to be their QA engineer I'll start taking the time to file bug reports.

That's a very scorched earth policy, do you have the same policy for the products or devices that you create?

"Hello yes the widget you designed then sold me broke, it's bad and you should feel bad. I could tell you how it broke but I need you to send me $100 first, then I will share the details with you."

[KE5FX]

If I think the vendor is making a good-faith effort to improve, then of course I'll work with them to help them understand the problem and fix it.

If I think the vendor is bumbling around breaking things that worked fine, doing so for no valid reason on God's green earth, I might bring a different attitude to the table.

[asmi]

I guess if you're passing high enough voltage through that cable it does not really matter what you use, so flash chips will do the trick. 

That's an MCU. But you point still stands

[gussy]

I made my first marketplace purchase through DigiKey this week. DigiKey sent out a quick 5 question, 1-5 stars survey after the order was received which makes me hopeful that persistently bad marketplace vendors will get kicked off the platform. We'll see how this goes longer term.

[james_s]

That's a very scorched earth policy, do you have the same policy for the products or devices that you create?

"Hello yes the widget you designed then sold me broke, it's bad and you should feel bad. I could tell you how it broke but I need you to send me $100 first, then I will share the details with you."

I'm already paying them for a service, and I see no indication that they are putting any effort into actually fixing the problems that people are complaining about so why should I waste my own free time finding their bugs for them? They should be paying testers to do this work for them so that the sort of bugs we're seeing do not escape into the field in the first place. My attitude about this sort of thing took it's current direction when Microsoft decided to outsource most of their QA to the consumer. The fact that they're too cheap to pay staff to perform that important task for them doesn't mean that I'm obligated to do it for them for free.

If I develop a widget I don't expect the customer to find all the bugs for me. Maybe there's an edge case that slipped through my processes in which case I'm happy to hear about it and I'll make an effort to fix it for them as soon as possible. But if I sell a widget that is a piece of crap with all kinds of glaring faults suggesting I couldn't be bothered to even test it at all before I sold it to the customer then I'd expect the customer to take their business elsewhere.

[tooki]

That's a very scorched earth policy, do you have the same policy for the products or devices that you create?

"Hello yes the widget you designed then sold me broke, it's bad and you should feel bad. I could tell you how it broke but I need you to send me $100 first, then I will share the details with you."

I'm already paying them for a service, and I see no indication that they are putting any effort into actually fixing the problems that people are complaining about so why should I waste my own free time finding their bugs for them? They should be paying testers to do this work for them so that the sort of bugs we're seeing do not escape into the field in the first place. My attitude about this sort of thing took it's current direction when Microsoft decided to outsource most of their QA to the consumer. The fact that they're too cheap to pay staff to perform that important task for them doesn't mean that I'm obligated to do it for them for free.

If I develop a widget I don't expect the customer to find all the bugs for me. Maybe there's an edge case that slipped through my processes in which case I'm happy to hear about it and I'll make an effort to fix it for them as soon as possible. But if I sell a widget that is a piece of crap with all kinds of glaring faults suggesting I couldn't be bothered to even test it at all before I sold it to the customer then I'd expect the customer to take their business elsewhere.

Your logic makes sense except for the fact that the premises it’s based on are entirely wrong. Microsoft never actually reduced testing: has more testers than programmers, and everything undergoes massive ongoing automated testing as well. (It’s the immense complexity of modern software (and hardware) that’s been responsible for a lot of the problems, as well as releases being driven by marketing decisions rather than doneness.)

It’s easy to think of others as being lazier than you, but in reality, that feeling is often unjustified.

[olkipukki]

DK is almost ready to became full-time "FBA" 

Digi-Key breaks ground on 2.2 million square foot Product Distribution Center expansion. First package from new facility is projected to ship in mid-late 2021.

http://www.digikey.com/en/resources/about-digikey#tabs-2

It looks like DK allows to sell anyone (of course who passed their onboarding team), not long time ago used to be only suppliers with USA warehouse...

Can I join the Digi-Key Marketplace if I don’t have a US-based warehouse?
Yes, you can!

 

[james_s]

Your logic makes sense except for the fact that the premises it’s based on are entirely wrong. Microsoft never actually reduced testing: has more testers than programmers, and everything undergoes massive ongoing automated testing as well. (It’s the immense complexity of modern software (and hardware) that’s been responsible for a lot of the problems, as well as releases being driven by marketing decisions rather than doneness.)

It’s easy to think of others as being lazier than you, but in reality, that feeling is often unjustified.

I feel that I have a better view into this than most observers since I worked at MS for about the first 10 years of my career as a professional tester during what I would consider their golden age in the Win98 through XP era when it was really a fantastic place to work, filled with some of the most amazing people I've ever met. I was actually there, I saw it all from the inside and don't have to speculate and although that was quite a long time ago I still have many good friends who work there. I left for greener pastures when they cut the pay scale for QA engineers roughly in half and made a big push to replace it with automation and developers unit testing their own code. Nadella quoted at some point "Who better to test the code than the developers who wrote it" to which I would say who WORSE to test it than the developers who wrote it, I could go on a whole rant detailing why but that doesn't really matter, just look at the market share for Windows 8 and the adoption percentage of the Edge browser as examples of the result. During that time friends still working there who were mostly developers and middle managers were complaining that they couldn't find decent testers, well, duh, they were trying to pay them peanuts while other companies still offered competitive wages. This led to a large talent bleed as experienced people fled to various startups, some of which later became big. For a while MS had almost no QA engineers at all but that started to change just a few years ago after several disastrous releases including data-loss bugs that escaped into the field that led to some major internal drama. Even now they still have the majority of their testing automated which is great for regression testing but automation will never see an edge case and go looking for it, and it's very, very easy to fall into a trap where the automation is passing so you think everything is fine, but it is rare that anyone really takes a close look at what sort of real world use cases the automation is covering. Being a good SDET is a lot like being a good technician, it's not as glamorous as being a SDE or hardware engineer but it requires a lot of the same skillset with an additional skillset on top of that. Developing useful automated test scripts is far more than coding, it requires the skills of a tester. They have their "insider program" or whatever it's called, which is absolutely an attempt to outsource testing to unpaid fanboys instead of paying internal staff to do it.

Anyway, bottom line is companies have paid me a very good salary (better than I've made in stints of doing EE related work) to find and document defects and deficiencies in their products, so forgive me if I'm not inclined to give away my services for free. We are discussing for-profit businesses here, not charities, and I don't think you will find many professionals in any field who are interested in working for free to help someone else earn money.